[SOLVED] web ui or ssh access stops working after a while

helcik

New Member
Sep 4, 2025
2
1
3
Hi,
Need to reboot the server to get access via web ui or ssh, then after couple minutes it looses management access.
I can still ping the server, can connect to VMs, just the WEB UI and ssh looses it. Did tcpdump on the interface for ssh and there's communication going but the client times out, same for https.

Running single node pve 8.4.12 on pretty recent hardware, tried changing IP, switching to my other eth port, same problem.

Any ideas how to troubleshoot that?
 
...then after couple minutes it looses management access.
I can still ping the server,

Are you sure the answer of "ping" came from this server? Or does it come from the machine using that IP address in that moment? (Hint: you can not know that without examining the MAC address.)

The first suspicion is an IP address conflict. I have seen exactly this behavior when one address was assigned to two machines.

I recommend strongly to configure all PVE nodes with a static address¹ and make sure that this one is outside of the DHCP range handled by your default DHCP server...


¹ https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_default_configuration_using_a_bridge
 
  • Like
Reactions: helcik and fba
With your advice I unplugged my server and client PC from the network switch and connected them directly with ethernet cable, I can see ping pockets coming in but server never replies back, same for ssh or https. Connected client PC back to the switch (server still unplugged) the server IP does not reply on ping, so there should be no IP conflict, I already tried different IP addressed in the past. Plugged server back into the switch, ping is coming back ok. Any advice what to try next?

Update:
Figured it out my interfaces file was wrong. I had

address 192.168.1.18/0

instead of

address 192.168.1.18/24


As a result it wouldn't add a proper route:
# ip route:
default via 192.168.1.1 dev vmbr0 proto kernel onlink

after fixing the file and restarting network all is good:

# ip route
default via 192.168.1.1 dev vmbr0 proto kernel onlink
192.168.1.0/24 dev vmbr0 proto kernel scope link src 192.168.1.18
 
Last edited:
  • Like
Reactions: UdoB
after fixing the file and restarting network all is good:
Great! I'm happy you get it going. You might click "Like" on my post if it helped...

Please tag this thread "Solved", there is an "Edit thread" button above the first post; left of the text is a drop-down menu :)